The Feeding, Larval D ispersal, and Metamorphosis of Philippia (: Architectonicidae ) '

ROB ERT ROB ERTSON, RUDOLF S. SCHELTEMA, and FRANK W . ADAMS2

ABSTRACT: In the Hawaiian Islands, Philippia (Psilaxis) radiata ( Reding) lives in sand or rubble near the hermatypic stony coral Porites lobata Dana, and emerges to feed at night on the polyps, Other species of Philippia (Psilaxis) probably have the same mode of life with corals. Philippia (Psilaxis ) veliger lar­ vae are abundant in tropical and subtropical oceanic plankton distant from any potential shallow-water hosts, and are dispersed great distances by near-surface currents. Duration of the pelagic larval stage is between several weeks and 6 months or longer. Metamorphosis, involving loss of the 4-10bed velum, initial growth of the teleoconch, and other changes, can precede contact with a host and is induced by capture from the plankton. Newly settled Philippia quickly attain a stage of arrested growth and can remain alive without feeding for several months . At this stage the postla rvae presumably crawl in search of hosts, and failure to find hosts doubtless causes the high mortality observed. Experiments at Woods H ole, Massachusetts, with newly metamorphosed P. (Psilaxis) krebsii (March), obtained as larvae from plankton in the Sargasso Sea, together with the ahermatypic coral Astrangia danae Agassiz, reveal physical prob­ lems for Philippia in assuming the adult mode of life with other corals. Young Philippia showed no ability to detect A strangia except by touch. Young Philippia lacked immuni ty to Astrangia nematocysts but were not seriously injured by them. The young gastropods are, however, subject to predation by this coral. Most con­ tacts with the living tissues of Astrangia caused a Philippia to be promptly drawn through the mouth and ultimately digested. The large protoconchs of Psilaxis would preclude their being swallowed by hermatypic corals such as Porites, with polyps smaller than those of A strangia. Proboscis eversion and feeding were not observed in young P. krebsii.

IN A PAPER reporting that the genus H eliacus ters. However, Nordsieck has informed us (in comprises obligate symbionts that feed on co­ litt.) that he has no more specific data than lonial zoanthiniarian sea anemones, Robertson that P. hybl'ida lives in a coral and coralline (1967) mentioned that nothing was known of habitat. Taylor (1968, p. 187) mentioned that the feeding habits of the two other major gen­ Arcbitectonica perspectioa (Linn.) "probably era in the Architectonicidae, Philippia and Ar­ feeds upon zoanthids" at 15 to 25 meters in cbitectonica. Subsequently, Nordsieck (196 8, the Seychelles, but Taylor has informed us (in p. 65) has mentioned the Mediterranean and lit!.) that he dredged A. perspectiua together Lusitanian species Philippia (Philippia) hy­ with zoanthids, and extrapolated from Robert­ brida (Linn.) "auf Koral len" at about 40 me- son's (1967) conclusions on Heliacus, Shuto (1969, p. 25, footnote) states that A. perspec­ 1 Woods Hole Oceanographic "Institution Contri ­ tiva is "strictly speaking . . . synbiotic [sic] bution No . 2312. Manuscript received May 9, 1969. with sand-dwelling sea anemones" ; no sup­ 2 Addr esses, respectively: Academy of N atural Sci­ porting data were given. Abbott (1968 , p. 86) ences of Phi ladelphia, Philadelphia, Pennsylvania mentioned Arcbitectonica nobilis Reding' "in 19103; Woods H ole Oceanographic Institution, Woods Hole, Massachusetts 02543; and 1739C Ala shallow water among sea pansies." Arcbitec­ Moana Boulevard, Honolulu, Hawaii 96815. tonica nobilis commonly is washed ashore on 55 56 PACIFIC SCIENCE, Vol. 24, January 1970

," .. . ,.. ...

,t) '~" ....-.... , i "2; \ >tt~ ~ ~~ ., ~'t:" ':"';'~ > : '.::~>:,.\~?:~.~ :. ' ,: .

5 mm .

FIG. 1. Postlarval Philippia (Psilaxis) radiata (Roding ) feeding on the polyps of the hermatypic stony coral Porites lobata Dana. Compos ite drawing by Mary Fuges from photographs of a Hawaiian feeding, from other photographs of extended Philippia , and from specimens of the shell and coral. The Pbilipp i« lives in the coarse sand near the coral and crawls out to feed at night. Philippia (Feeding, etc.) ---':"'ROBERTSON ET AL. 57

Texas beaches together with pennatulids Our information on larval dispersal and the (Renilla muelleri Kolliker; teste R. T. Abbott) . metamorphosis of Philippia enhance the in­ Most recently, A. nobilis has several times been terest of the nutritional dependence of the post­ seen with its proboscis everted, presumably larvae on coelenterates. All three species in the while feeding on something (Futch, 1969). subgenus Psilaxis have planktotrophic veligers Thus, all five of these tantalizing reports pro­ that remain for long periods in the plankton vide no definite information. There are simi­ (Robertson, 1964 ; Scheltema, 1968; Robert­ larly vague hints that Gyriscas, a fourth archi­ son, 1970b) . We have but scanty data on the tectonicid genus, may live with gorgonians two Indo-Pacific species, and therefore we sum­ (Tiberi, 1868) and sponges (Powell, 1965, marize here some of our data on the larvae of pp. 161-162). the Atlantic Ocean species Philippia (Psilaxis) New observations (by Adams) are here re­ erebsii (March) , on its metamorphosis in the ported on Philippia (Psilaxis) radiata (Ro­ laboratory at Woods Hole, and on the inter­ ding) which in the Hawaiian Islands lives action between the young postlarvae and the near and feeds on the soft tissues of the herma­ ahermatypic coral Astrangia danae Agassiz. typic stony coral Porites lobata Dana (Fig. 1). Information on the systematics, zoogeography, (This discovery was briefly mentioned in Ha­ and bathymetry of Indo-Pacific Philippia (Psi­ waiian Shell News, vol. 15, no. 10, p. 4; Oc­ laxis) , and on eggs and young postlarvae in tober 1967) . The Architectonicidae become the umbilicus of adults, is given in the follow­ the fifth family of prosobranch gastropods ing paper (Robertson, 1970b) . definitely known to include predators or para­ sites on stony (scleractinian) corals (Robert­ PROBOSCIS STRUCTURE son, 1970a). The habitat and feeding data are preceded here by a review of the structure of Like other architectonicids, Philippia has a the proboscis, jaws, and radula of Philippia. remarkably long, acrembolic proboscis. When Their specialization and uniformity lead us to this is fully everted, the true mouth and the believe that food and mode of feeding are uni­ small buccal mass containing the radula are at form throughout the genus. the tip. Thus the animal can feed only when

FIGS . 2 and 3. Shell of the PhiZippia radiat a specimen seen to have fed on Porites Zobata (45 feet off Makua, western Oahu, Hawaiian Islands ; shell in the collection of Frank W . Adams) . Apical and basal views, respect ively (both X 6) . 58 PACIFIC SCIENCE, Vol. 24, January 1970

the proboscis is fully everted. The lining of the genetic relationship and by Thiele (1928, pp . buccal cavity and esophagus is cuticularized as 87- 88) to convergence. Epitoniids are now in epitonii ds, and this prevents injury from known to feed on coelenterates exclusively and nematocysts in both groups (Robertson, 1970b). we are inclined to agree with Thiele. The laterally paired jaws of Philippia are elon­ gate, are comprised of scalelike elements, and Philipp ia radiata: are the distal part of the buccal cavity wall. HAWAIIAN HABITAT AND F EEDING DATA Upon complete proboscis eversion, the jaws are near the mouth edge. The radula, uniform Philippia (Psilaxis) radiata is one of two throughout the genus, is narrow and comprised 'closely related species in the H awaiian Islands of elonga te teeth bent distally and with one to (Robertson, 1970b) . The postlarvae, rarely six points. As observed long ago by T roschel collected alive, are known from shallow water (1875, p. 155), these architectonicid radu lar to about 150 feet. During daylight, the ani­ teeth resemble those of epitoniids. The resem­ mals are usually buried 1 to 4 inches in silty blance was attributed by Troschel to phylo- sand or loose, algal-covered rubble, the sand

FIG. 4. Porites /obata. Some of the calices of the cleaned skeleton of the coral seen to have been fed on by Pbilippi« radiate. X 30. Philippia (Feeding, etc.) - ROBERTSON ET AL. 59 and rubbl e shallowly overlying basalt (Ha­ boscis fully everted adjacent to the Porites col­ waiian Shell N ews, vol. 16, no. 12, p. 7; ony as it fed delicately on the still partly December 1968) . In every case known to us, expanded polyps. Under bright photoflood the living animals were near massive or incrust­ lights, two color photographs were taken be­ ing colonies of Porites lobata. fore the Philipp ia inverted its proboscis and One living Philippia radlata with a fairly retracted into its shell. Th e drawing (F ig. 1), small shell ( Figs. 2 and 3) was collected by based in part on these photographs, probably Adams while scuba diving at a depth of 45 does not show the full length of the extended feet off Makua, western Oahu, in April or proboscis. May 1967. It was taken from a patch of sand near a small living colony of Porites lobata on SPECIFICITY OF Philippia TO CORAL HOSTS a mainly dead skeleton resting on the sand. o Both the gastropod and the coral were kept Extrapolating from the Hawaiian habitat alive in an aquarium. The Philippia remained data, from the single feeding observation, from alive for about two months and was secretive, the fact that the proboscis is adapted for feed­ retracting into its shell whenever ordinary light ing on coelenterates, and from the much more was shone on it. Infrared light was found not thorough knowledge of the mode of life of to disturb the animal. Late one evening, under H eliacus, it seems probable that Philippia is infrared, the Philippia was seen with its pro- host-specific to stony corals. The world distri-

FIG. 5. Incompletely developed veliger larva of Pbilip pia (Psi/axis) erebsi) (March) with the 4-lobed velum full y exten ded, show ing also the operculum and non-tentaculate eyes. From 38°08' N , 69° 55' W . X 50. The fully developed larva has longer velar lobes. 60 PACIFIC SCIENCE, Vol. 24, January 1970 bution of Psilaxis coincides closely with areas stages in the collecting program, see Scheltema, where there are more than two to five genera 1966 and 1968). The larvae live in the same of hermatypic corals (Steh li, 1968, p. 209, fig. depth range as the adults but freq uently are 45) . We infer that postlarval Philippia (Psi­ distant from any potential hosts. The known laxis) is obligately associated (symbiotic) stations of occurrence range from latitud es with hermatypic corals, and that the soft tissues 44°N (east of Nova Scotia) to 18°5 (east and mucus of corals are its exclusive food . of Brazil) and from longitudes 85°W (west­ The coral on which our feeding observations ern Caribbean) to 8oW (south of Liberia) . were made (Fig. 4) was identified for us by This area includes 9.1 million square nautical Professor John W . Wells as Porites lobata miles (27 million square kilometers) of ocean. Dana. Among the other symbionts with corals, Out of a total of 502 plankton samples from host-specificity to particular genera or families different stations within the area of occurrence, is uncommon (Robertson, 1970a) . There is no larval P. krebsii were foun d in 113 ( 22 per ­ reason to suppose that Philippia radiata is spe­ cent of the total) . Sea surface temperatures cific to Porites lobata. N everth eless, Porites lo­ ranged from 29°C down to 19°C, but most bata is potentially a major host for Philippia of the larvae probably were at depths below radiata in the tropical west and central Pacific the warmest water. because it is abundant in the Hawaiian Islands (Vaughan, 1907, pp . 196-207) and wide­ Duration of the Pelagic Larval Stage spread elsewhere, being known also from the The maximum duration of the pelagic larval Marshall, Line, Samoa, and Fiji islands, and stage, deduced from the localities of occur­ from the Great Barrier Reef, Australia (Cross­ rence and the maximum current velocities from land, 1952, p. 242 ; Wells, 1954, p. 452). the nearest potential spawning areas, is in the Porites lobata is not known from the Indian order of six months or longer. Trans-Atlantic Ocean, and so Philippia radiata must have dif­ transport by ocean currents must be frequent, ferent hosts in this ocean. although neither larvae nor adults are yet known from the W est African continental shelf." LARVAL AND POSTLARVAL Philippia krebsii The minimum duration of the pelagic stage Larval Dispersal is of interest because this probably is critical in maintaining local populations. Incompletely As has been reported elsewhere (Robertson, grown veligers of Philippia krebsii occur no 1964; Scheltema, 1968, p. 1160, fig. 1), the farther from spawning areas than ocean cur­ pelagic veliger larvae of Philippia (Psilaxis) rents could transport them within several krebsii are remarkably abundant in near-sur­ weeks. Thus, growth of the veliger is rapid, face waters of the trop ical and subtropical full size is quickly attained, and settlement Atlantic Ocean. So far, the benthic adults would seem possible within several weeks (fairly scarce in museums) are known only after hatching. W ith hydrographic conditions from the western Atlantic (Barbados, Curacao such as those at Bikini (Johnson, 1949) and and Yucatan nort h to Bermuda and off Cape Bermuda (Boden, 1952; Boden and Kampa, Hatteras ) and from two areas in the eastern 1953), larval Philipp ia (Psilaxis) could some­ Atlantic (Canary and Cape Verde Islands). times settle near their parents. The geographic The known bathymetric range is 4 to 112 fath­ variations in pro toconch size ( Robertson, oms (7 to 205 meters), the living animals 1970b) are an indication that some of the being mostly in the shallower dept hs. populations are truly semi-isolated . Plankton collections made from various re­ search vessels during the past several years 3 In a paper on coral- inhabiting crabs in the east­ have revealed that larval P. krebsii occurs the ern Pacific, Garth and H opkins ( 1% 8, pp . 45-46), entire distance across the Atlantic, offshore quoting W . K. Patton, argue that highly adapted "commensals" with corals are more likely than are over deep water as well as in coastal regions. free-living species to establish populations after tran s­ ( For methods and other results from earlier oceanic tra nsp ort , Philippia (Feeding, etc.)-ROBERTSON ET AL. 61

Living Larvae amber at the sutures, on the anal keel, and The new observations here reported on lar­ on the left side near the varix ) . The four val and young postlarval Philippia erebsii are velar lobes ( Fig. 5) lengthen as the larva based on living animals from plankton in the grows. There are paired eyes but no tentacles Gulf Stream and in the western Sargasso Sea. until metamorphosis. Probably on account of Repeatedly, these were brought for study to the green pigmentation of the unicellular algal Woods Hole. Here they frequently survived food, the pale brown digestive gland has a for several months in small containers of Mas­ greenish cast. Seen in transparency through the sachusetts sea water changed daily or weekly, shell on the right side of the body is a con­ and with or without added food (unialgal cul­ spicuous (opisthobranch-like) black larval ex­ ture) . A detailed description of the morphol­ cretory organ that is irregularly oval in outline. ogy and coloration of the veliger is beyond the scope of this paper, and will be included Metamorphosis in the Laboratory in another study in progress on the larvae of Upon capture from the plankton, the larvae all Atlantic Ocean achitectonicids. invariably lack any growth of the teleoconch The hyperstrophically coiled larval shell is (postlarval shell). If they are fully developed smooth and lustrous, and before metamorphosis (i.e., if the peritreme is thickened into a it is transparent and mainly colorless (pale varix) , the larvae lose their velums within a

F IG. 6. You ng post larval Philippia krebsii in the stage of arrest ed growth, showing the smoo th and lustrous protoconch (l arval shell ) wit h its consistent color pattern that develops du ring metamorph osis, the teleoconch (postl arval shell) with its brown spiral cords crested with periostracum, the opa que wh ite pigm ent spots on the translucent white body, the broad foot with its median anterior cleft, the pair of tentacl es, the right eye positioned on the later al base of the tentacle, the round operculum on the hind end of the foot (seen in an obliq ue view) , and (i n transparency thr ough the shell) the elongate, vesicular, black larval excretory orga n. Larva from 35° 27' N , 66 °53' W .X 50. 62 PACIFIC SCIENCE, Vol. 24, January 1970

few days. Settlement in the laboratory occurs about on its broad foot (Fig. 6) . After devel­ in the absence of corals, and apparently is in­ oping from between one-sixth to about one­ duced by contact with any solid surf ace. half a whorl in one or two days, growth of At metamorphosis, the velum disappears the teleoconch stops. Animals remained alive first; the ciliated edges drop off in fragments in this state of arrested growth without fur­ and the reminder is resorbed . Subsequently, ther changes for several months. The mini­ the digestive gland loses its greenish cast, and mum temperature in the laboratory was 20.5°C. slowly becomes more granular exteriorly. The protoconch darkens- especially on the right Postlaroae with Astrangia danae side, where the consistent brown and white Knowledge that postlarval Philippia feeds pattern appears. Th en the black larval excretory on the polyps of stony corals prompted ex­ organ slowly length ens and becomes vesicular. periments at W oods Hole with young post­ N ext, tentacles appear and the body assumes larval P. krebsii to see if they would feed on its postlarval coloration (translucent white, Astrangia danae, the only local stony coral. with opaque white spots on the tentacles and Unl ike Porites, Astrangia is a nontropical, upp er side of the foot) . About a week after ahermatypic coral with large polyps. The re­ capture, the teleoconch suddenly begins to sults were negative in that young Philippia grow, and the animal then actively crawls krebsii was not seen even to evert its proboscis.

FIG. 7. Young postlarval Philippia krebsii being drawn through the large mouth of a polyp of the ahermatypic stony cora l A strangia danae Agassiz . X 30. PhiJippia (Feeding, etc.)-RoBERTSON ET AL . 63

The observations reveal problems for Pbilippia family there is the same sequence of structural inherent in assuming an adult mode of life and behavioral changes at settlement, and that with corals. the stage of arrested growth seen in the labora­ When a young PhiJippia was placed on the tory is normal. At this stage the animals pre­ broad oral disc of Astrangia, a feeding re­ sumably crawl .in search of their hosts. Many sponse was elicited not in the gastropod but architectonicids die at this stage: the Recent in the coral, which clasped the shell with the and fossil shells are common in museum col­ adhesive nematocysts on its tentacles and lections.! We suggest that this high mortality quickly drew the gastropod through its mouth is caused mainly by the spatial problems in into the gastrovascular cavity (Fig. 7) . One finding hosts at this critical stage in the life Pbilippia survived several hours in an Astran­ cycle. gia gastrovascular cavity without apparent in­ jury, but another was completely digested in CONCLUSIONS less than 18 hours (the empty shell and oper­ culum were expelled afterwards). The life cycle of PhiJippia (PsiJaxis) is re­ The behavior of Pbilippia near Astrangia markable because the larvae and young post­ was best seen at night; as with PhiJippia radi­ larvae have to overcome major spatial and ata, P. krebsii is most active then. Young P. physical problems in order to maintain the krebsii showed no ability to detect Astrangia populations of adults . The long-lived larvae polyps except by contact. Whenever the ten­ are spawned in prodigious numbers and are tacles or anterior end of the foot of a Philip­ dispersed over enormous areas and distances pia touched any part of the living tissues of by near-surface ocean currents. To grow and Astrangia, the gastropod reacted by quickly attain sexual maturity, the postlarvae must first withdrawing into its shell and expelling mucus. find, and then live near, shallow subtidal coral The coral usually reacted by swallowing the polyps-presumably their exclusive food . The PhiJippia. The gastropod lacked immunity to larvae frequently are distant from any poten­ Astrangia nematocysts but seemed not to be tial hosts, over deep water and in shallow water seriously injured by them, recovering within where corals are sparse or absent. Larval mor­ several hours from extensive contact. tality is extensive. However, the minimum du­ The corals normally fed upon by PhiJippia ration of the pelagic stage may enable some of (Psilaxis) probably cannot swallow the young the larvae to settle near their parents. postlarval gastropods. The calices, polyps, and Settlement can precede contact with a host, mouths of Astrangia are larger than those of and in metamorphosis the velum is lost first. Porites (Figs. 4 and 7, at the same scale). The This suggests that the swimming-crawling maximum diameter of the elastic mouth of (pediveliger) stage is very short, and that the Astrangia is about 2 mm. This is barely large young postlarvae (in a state of arrested enough to allow passage, and a Porites mouth growth) crawl in search of hosts. Mortality is would be too small. extensive at this stage also. The young appear to have no means of detecting coral polyps EARLY POSTLARVAL ARRESTED GRO~TH except by touch, and they lack immunity to nematocysts. In the laboratory they are also The stage of early postlarval arrested growth subject to predation by large coral polyps, but is not confined to PhiJippia but is recorded on probably not by their normal hermatypic hosts. well-preserved shells of all postlarval archi­ tectonicids. Invariably there is one (rarely two 4 Such shells are figured by Robertson ( 1964, or more) distinct growth lines in the first or figs. 12-17 [Phi/;pp ia]) , Hedley (1903, pp . 349­ second quadrant of the first teleoconch whorl 350, fig. 73 [Architectollica]), and Janssen (1967, (Woodring, 1959, pp. 168-170, pl. 30, fig. 7; p. 133, pl. 7, figs. 1-2, 4 [vari ous fossil architec­ Robertson, 1964, fig. 4; Jung, 1969, p. 455, tonicidsJ ). Janssen stud ied a collection from a Mio ­ cene faun a in Germany; there were six species and pl. 46, fig. 4; Robertson, 1970b, figs. 1-2). 149 · specimens, 24 (1 6 percent) of wh ich wer e j u­ This uniformity suggests that throughout the veniles . 64 PACIFIC SCIEN CE, Vol. 24, January 1970

ACKNOWLEDGMENTS luskenfauna 1. Geologica et Palaeontologica [Philipps-Universitat, Marburg] , vol. 1, pp. This work was supported by National Sci­ 115- 173, 8 figs., 14 pIs. ence Foundation Grants GB-7008, GB-2207, JOHNSON, M. W . 1949.Zooplankton as an and GB-6027X, and was done in part while index of water exchange between Bikini la­ Robertson was a Guest Investigator at the goon and the open sea. Transactions of the Woods Hole Oceanographic Institution. Most American Geophysical Union, vol. 30, pp . of the larvae were obtained by Scheltema dur­ 238-244, 15 figs. ing cruises of the research vessels "Atlantis JUNG, P. 1969. Miocene and Pliocene mol­ II," "Chain," and "Gosnold." We thank Pro­ lusks from Trinidad. Bulletins of American fessor John W . Wells (Cornell University, Paleontology, vol. 55, pp . 293-657, 4 figs., Ithaca, New York) for identifying the Porites, pIs. 13-60. and Mary Fuges for drawing Figure 1. NORDSIECK, F. 1968. Die europaische n Meeres­ Gehauseschnecken (Prosobranchia) vom Eis­ LITE RATURE CITED meer bis Kapverden und Mittelmeer. Fischer, Stuttgart. viii + 273 pp., 16 color figs., 31 ABBOTT, R. T. 1968.Seashells of No rth Amer­ pIs. ica; a guide to field identification. Golden POWELL, A. W . B. 1965.N ew Zealand mol­ Press, New York. 280 pp ., many figs. luscan systematics with descriptions of new BODEN, B. P. 1952. Natural conservation of species: part 5.Records of the Auckland insular plankto n. Nature, vol. 169, pp . 697­ Institute and Museum, vol. 6, pp. 161-168, 699, 2 figs. pIs. 22-23. BODEN, B. P., and E. M. KAMPA. 1953. W in­ ROBERTSON, R. 1964. Dispersal and wastage ter cascading from an oceanic island and its of larval Philippia krebsii (Gas tropoda: biological implications. Nature, vol. 171, pp. Architectonicidae) in the N orth Atlantic. 426-427, 2 figs. Proceedings of the Academy of Natural CROSSLAND, C. 1952. Madreporaria, Hydro­ Sciences of Philadelphia, vol. 116, 27 pp., corallinae, H eliopora and Tubipora. British 6 tables, 17 figs. Museum (Natural History) Great Barrier --- 1967. H eliacns (Gastropoda : Archi­ Reef Expedition, Scientific Reports, vol. 6, tectonicidae) symbiotic with Zoanth iniaria pp . 85-257, 1 fig., 56 pIs. (Coelenterata) .Science, vol. 156, pp . 246­ FUTCH, L. 1969. Observations on Arcbitec­ 248, 1 table, 1 fig. tonica nobilis. Miami Malacological Society - -- 1970a. Review of the predators and Quarterly, vol. 3, no. 1, 2 pp. parasites of stony corals, with special refer­ GARTH, J. S., and T. S. HOPKINS. 1968 . Pse11­ ence to symbiotic prosobranch gastropods. docfyptochiflls crescentus (Edmondson), a Pacific Science, vol. 24, pp. 43- 54. second crab of the corallicolous fam ily - -- 1970b. Systematics of Indo-Pacific Hapalocarcinidae (C rustacea, Decapoda) Philippia (Psilaxis) , architectonicid gastro­ from the eastern Pacific with remarks on pods with eggs and young in the umbilicus. phragmosis, host specificity, and distribution. Pacific Science, vol. 24, pp . 66-83, 17 figs., Bulletin of the Southern California Academy 1 table. of Science, vol. 67, pp. 40-48, 2 figs. SCHELTEMA, R. S. 1966. Evidence for trans­ H EDLEY, C. 1903. , Part II . Scaphop­ Atlantic transport of gastropod larvae be­ oda and Gastropoda. Scientific results of longing to the genus Cymaiium . Deep-Sea the trawling expedition of H.M.C.S. "Thetis" Research, vol. 13, pp . 83-95, 2 tables, 5 figs. off the coast of New South W ales. . --- 1968. Dispersal of larvae by equatorial Australian Museum Memoirs, vol. 4, · pp. ocean currents and its importance to the zoo­ 325-402, figs. 61- 113, pls, 36-38. geography of shoal-water tropical species. JANSSEN, A. W . 1967. Beitrage zur Kenntnis Nature, vol. 217, pp . 1159-1162,4 figs. des Miocans von Dingden und seiner Mol- SHUTO, T. 1969. Neogene gastropods from Philippia (Feeding, etc.) - ROBERTSON ET AL. 65

Panay Island, the Phil ippines. Memoirs of testace de la Mediterranee, Journ al de Con­ the Faculty of Science, Kyushu University, chyliologie, vol. 16, pp. 56-60, pI. 5, fig. 1. ser. D (Geology), vol. 19, no. 1, 250 pp. , TROSCHEL, F. H. 1875. Das Gebiss der 43 figs., 5 tables, 24 pIs. Schnecken.Vol. 2. Berlin. Pp. 133- 180, pIs. STEHLI, F. G. 1968. Taxonomic diversity gra­ 13-16. dients in pole location : the Recent model. VAUGHAN, T. W . 1907. Recent Madreporaria In : E. T. Drake, ed., Evolution and environ­ of the H awaiian Islands and Laysan. U. S. ment, pp. 163-227, 56 figs. Yale University National Museum Bulletin 59, ix + 427 Press, New Haven and London. pp. , 1 map, 96 pIs. TAYLOR, J. D. 1968. Coral reef and associated WELLS, J. W . 1954. Recent corals of the Mar­ invertebrate communities (mainly molluscan) shall Islands. U. S. Geological Survey Pro­ around Mahe, Seychelles. Philosophical fessional Paper 260-1, pp. i-iv, 385-486, 4 Transactions of the Royal Society of London, tables, figs. 119-122, pIs. 94-187. ser. B, vol. 254, pp . 129-206, 20 figs., pIs. WOODRING, W. P. 1959. Geology and paleon­ 13-17. tology of Canal Zone and adjoining parts of THIEL E, J. 1928. Dber ptenoglosse Schnecken. Panama; description of Tertiary mollusks Zeitschrift fur W issenschaftliche Zoologie, (gastropods:Vermetidae to Thaididae). U. S. vol. 132, pp. 73-94, 11 figs. Geological Survey Professional Paper 306-B, T IB ERI, N. 1868. Sur un nouveau genre de pp. i- iii, 147-239, pIs. 24--38.